SQL Server VAR() Function

The SQL Server (Transact-SQL) VAR() function returns the statistical variance of all values in the specified expression.

Syntax

The syntax for using VAR() function is given below:

SELECT VAR(column_name) 
FROM table_name
WHERE condition(s);

The SQL Server (Transact-SQL) GROUP BY clause is used to arrange result table into identical groups when one or more columns are used. Please note that it is must to include those column names in a GROUP BY clause which are not encapsulated within the VAR() function. See the syntax below:

SELECT column1, column2, ...
       VAR(column_name) 
FROM table_name
WHERE condition(s)
GROUP BY column1, column2, ...;

Parameters

Return Value

Returns the statistical variance of all values in the specified expression.

Example - With Single Column

Consider a database table called Employee with the following records:

To get the variance of salary of all employees whose age is greater than 25, the following query can be used:

SELECT VAR(Salary) AS VAR_Salary 
FROM Employee
WHERE Age > 25;

This will produce the result as shown below:

Example - Using DISTINCT

The DISTICT clause can be used with VAR() function. For example - To get the variance of distinct (unique) salaries, the following code can be used:

SELECT VAR(DISTINCT Salary) AS VAR_UniqueSalary 
FROM Employee;

This will produce the result as shown below:

Example - Using Formula

The expression contained within the VAR() function does not need to be a single field. A formula can also be used with this function. For example - Consider a bonus (15% of Salary) is given to each employee, and to calculate the variance of bonus amount, the following query can be used:

SELECT VAR(Salary * 0.15) AS VAR_BonusAmount
FROM Employee;

This will produce the result as shown below:

Example - Using GROUP BY

To get the variance of salary of employees group by their age, the following query can be used:

SELECT Age, VAR(Salary) AS VAR_SalaryByAge 
FROM Employee
GROUP BY Age;

This result of the above code will be: